1953 Gibson GA-40 Tremolo

[Tom Phillips]

Hi All,


I have a 1953 Gibson GA-40 on the bench. This is the original Les Paul amp with octal pre-amp tubes.


Everything is now working but the tremolo sound is weak and choppy. When I scope the output and adjust to display the modulation envelope I see that the modulation depth is a little less than 50% and the duty cycle is much less than 50%. The envelope shape is pretty squared off. Not at all a nice sine shape as achieved by Fender output tube bias modulation.


Anyway, my question to the group is:
Has anyone heard one of these early 1050s GA-40 amps that had good rich sounding tremolo? I ask because the circuit checks out OK and I don’t want to try to improve something that is already working as well as it ever was. (I’ve chased that Star before and I already have enough hours in this project) I have heard good things about the rich sound of this model amp but I have never heard anything about the tremolo quality.


This is the only amp I have seen that modulates the screen voltage of the output tubes screen to create the tremolo. Note that the 1953-54 model GA-40 circuit is completely different than that of the later versions of the same model number.


I plan to keep the amp stock for the customer. It would be very helpful if someone can share some first hand experience with the tremolo for this exact model amp. Then I’ll either look deeper or call it done.


Thanks in advance,
Tom

[wallyo]

The tremolo oscillator is a text-book phase-shift oscillator so if its waveform is not sinusoidal I would suspect that one or more of the caps (the 0.05 ones) is either off-value or leaky or both. Check the waveform at the plate or the grid of the 6SQ7.
I've forgotten some of the details about this type of oscillator but I know the gain has to be >29 to overcome the loss of the phase shift network. I believe that since they have chosen to vary only one of the resistors the gain has to be greater and the waveform will have more distortion.
If the gain is way too high I'm thinking that might lead to even more distortion but having said that I don't think the circuit they used can go very much higher than maybe 35 or 40.

[octal]

That is definitely a unique trem circuit. Are you sure the trem works by modulating the screens and not by just forcing the cathodes of the output tubes up and down? (i.e. bias modulation via the cathodes.) I honestly don't know. It's a confusing circuit. Given the low parts count, it should be possible to check and or substitute all the components in the trem circuit. That cap that shunts across L2 is probably pretty important to how the circuit operates. Just some seat of the pants electronic intuition here, I'm not claiming to be an expert on this circuit.

[52 Bill]

Is this the model with the three 6V6s and the three position speed switch? I just had one on the bench and I though that the trem was nice enough, but a little weak in the depth department.